I really don't see where the need to overclock comes from anymore. Today's speeds are pretty darn fast and I'd assume that if you actually have a real need for more processing power, that you should be able to come up with the couple hundred bucks for another socket/proc. Or uhh, overclock it?
Lately I've been undervolting to build silent systems. The latest AMD Brisbane processors at 2.1GHz can be undervolted to 1.05V and still pass my stress tests at speed, and stay below 40C with the 'silent' fan modes. Okay, now we see your true motivation for the above.
Anyways - as I'm sure you've experienced - we have the ability to undervolt AND hit peak voltage at the same time (with a base voltage set in BIOS of course).
I always find it amusing the irony of the underclocker fanboys. As if hitting the lowest possible temperature is manifest destiny. You realize that typical AMDs are stable to around 75C right? You also realize that a CPU is a solid state chip and not a car alternator as well?
Yes I realize it can "save money" through lower voltage used, but it's splitting hairs these days, and voltage scaling still wins out.
One thing to note: remember the days of before with CPU locks and other various schemes to prevent overclocking by the owner. Now it seems like we've come full circle.
One thing I do continually find funny is the constant disclaimer of "fried" CPUs. Anyone who overclocks a CPU knows it's fairly difficult to actually damage it unless one is doing something incredibly stupid.
At one point in time - they were, or atleast were able to successfully operate as so (pre-97/98 or so). The onset of internet popularity in the late 90s coupled with the need for lawmakers to pin legal responsibility on *someone* resulted in common carrier rights (even if presumed) of some ISPs ending up being yanked from them.
So if the producer compresses stuff 3:1, and the mastering engineer compresses stuff 4:1, the net result is 12:1, not 7:1. While I agree with pretty much the rest of your post, this last part isn't exactly accurate as it doesn't take into account comp threshold. Compression, as I'm sure you know, is definitely not just ratio and typically higher ratios or "harder" compression are applied against higher threshold settings and hence operate on a more narrow "band" of dynamics. In essence I wouldn't be surprised if the resultant ratio in your example is actually below 7:1, on average, after everything is said and done with.
Actually I am. I'm doing their brain a favour by not distracting them with prose-like variable names and instead being clear and concise with what matters: succinct identifiers that don't pander to insecure programmers.
" we're also working to increase the capabilities of those apps by adding functionality to the Web itself, such as providing support for offline data storage and access to 3D graphics hardware.'""
Look up interdict in Wiktionary, the free dictionary. In law, an interdict is a court order enforcing or, more commonly, prohibiting a certain action; a synonym for "injunction". Types of interdicts include: a restraining order (where, for instance a person is prevented from threatening, harassing, or abusing another person); a suppression or "gagging" order (where, for instance, a newspaper is prevented from making certain information public); and an enforcement order (where, for instance, a company is forced to withhold wages from a person who has failed to meet maintenance or other financial obligations). In the Roman Catholic Church, interdict is an ecclesiastical penalty which (temporarily) bars a specific person or group of people from receiving the sacraments. A type of shield or military defense against attack. Conversely, the act of attacking or interrupting a military logistics system, for example air interdiction. In Roman Law, it was (under the name interdictum) an order of a "praetor"
--
Anyways, these guys are toast. There can be no recovery.
I'm going to call bullshit on most of this, except that dysfunct's post seems the one with the most actual insight - that being the point that not all processes demand CPU as 100% of their running time.
Not all aformentioned OS' even calculate load using the exact same algorithm. Some may choose to include processes waiting on I/O and some may include only processes waiting on CPU.
I've personally had a Solaris 9 box up to a load of 1100+ doing testing with thread capacity handling, etc. and while the host took around 20 seconds to even handle my SIGINT to stop the main load generating process. While I highly doubt anything interactive would function fluidly, I pretty much interpreted the load average as a rough indicator rather than anything concrete.
However, Steinberg's use of the particular dongle method and provider (http://www.syncrosoft.com) took the last group 1000s of man hours to crack and they vowed to not even bother with the next release (the actual group was "H2O" and they provided a small background summary in one of the Cubase release NFOs). So in reference to the OP's question? Well, that far, if he wants it to stay uncracked for a decent amount of time.
Otherwise, much like an average size vs space analogy, it's convience vs security in this context.
Slashdot Mirror
Anyways - as I'm sure you've experienced - we have the ability to undervolt AND hit peak voltage at the same time (with a base voltage set in BIOS of course).
http://cpu.rightmark.org/products/rmclock.shtml
CnQ, SpeedStep, etc.
I always find it amusing the irony of the underclocker fanboys. As if hitting the lowest possible temperature is manifest destiny. You realize that typical AMDs are stable to around 75C right? You also realize that a CPU is a solid state chip and not a car alternator as well?
Yes I realize it can "save money" through lower voltage used, but it's splitting hairs these days, and voltage scaling still wins out.
That guy wishes he was Tommi Makinen!
One thing to note: remember the days of before with CPU locks and other various schemes to prevent overclocking by the owner. Now it seems like we've come full circle.
Even Intel is embracing it.
One thing I do continually find funny is the constant disclaimer of "fried" CPUs. Anyone who overclocks a CPU knows it's fairly difficult to actually damage it unless one is doing something incredibly stupid.
"have never been" is not accurate.
At one point in time - they were, or atleast were able to successfully operate as so (pre-97/98 or so).
The onset of internet popularity in the late 90s coupled with the need for lawmakers to pin legal responsibility on *someone* resulted in common carrier rights (even if presumed) of some ISPs ending up being yanked from them.
This was a fairly known case:
http://www.loundy.com/CASES/RTC_v_Netcom.html
http://rtfm.insomnia.org/~qg/trolls-need-attention.php
Based on the relative "robustness" of past DRM-like encryption schemes - I expect this to be voraciously cracked within a couple of days upon release.
The people impementing said encryption schemes are not the same individuals cracking them. Because of that, the schemes will be continually cracked.
Think: narcotics officer whose never done drugs.
But with the latest testimony by Rory Reiser - it really doesn't look good for Hans:
http://blog.wired.com/27bstroke6/hans_reiser_trial/index.html
So I can stop having to scroll past these Google and Microsoft articles.
Go USA!
"Stolen", my ass.
Then why use them as an example to back up 44 vs 48? It's more so the quality of said gear, in your example, than 44 vs 48.
Amazing. So powerful - the concept has been part of Unix for decades now.
Way to go MS - keep innovating!
Actually I am. I'm doing their brain a favour by not distracting them with prose-like variable names and instead being clear and concise with what matters: succinct identifiers that don't pander to insecure programmers.
" we're also working to increase the capabilities of those apps by adding functionality to the Web itself, such as providing support for offline data storage and access to 3D graphics hardware.'""
forest for the trees.
/* please don't use MixedCase bullshit */
Dude. Seriously consider the legal ramifications of hosting that.
This comment is on the money.
Well you've got the gmail password, could you go check out those latest forwards for us? Thanks.
I can see it bothers you quite a lot.
Flex those muscles dude. You feel strong? Now get in the back and install that new Microsoft Office update, thanks.
From wikipedia.
Interdict can refer to several things:
Look up interdict in Wiktionary, the free dictionary.
In law, an interdict is a court order enforcing or, more commonly, prohibiting a certain action; a synonym for "injunction". Types of interdicts include: a restraining order (where, for instance a person is prevented from threatening, harassing, or abusing another person); a suppression or "gagging" order (where, for instance, a newspaper is prevented from making certain information public); and an enforcement order (where, for instance, a company is forced to withhold wages from a person who has failed to meet maintenance or other financial obligations).
In the Roman Catholic Church, interdict is an ecclesiastical penalty which (temporarily) bars a specific person or group of people from receiving the sacraments.
A type of shield or military defense against attack.
Conversely, the act of attacking or interrupting a military logistics system, for example air interdiction.
In Roman Law, it was (under the name interdictum) an order of a "praetor"
--
Anyways, these guys are toast. There can be no recovery.
Nope, no global warming here Bob...
Bucket of water.
Apparently you haven't been using unix that long have you? -20-to-20 has been around for ages and is by no means relegated to Linux.
I'm going to call bullshit on most of this, except that dysfunct's post seems the one with the most actual insight - that being the point that not all processes demand CPU as 100% of their running time.
Not all aformentioned OS' even calculate load using the exact same algorithm. Some may choose to include processes waiting on I/O and some may include only processes waiting on CPU.
I've personally had a Solaris 9 box up to a load of 1100+ doing testing with thread capacity handling, etc. and while the host took around 20 seconds to even handle my SIGINT to stop the main load generating process. While I highly doubt anything interactive would function fluidly, I pretty much interpreted the load average as a rough indicator rather than anything concrete.
However, Steinberg's use of the particular dongle method and provider (http://www.syncrosoft.com) took the last group 1000s of man hours to crack and they vowed to not even bother with the next release (the actual group was "H2O" and they provided a small background summary in one of the Cubase release NFOs). So in reference to the OP's question? Well, that far, if he wants it to stay uncracked for a decent amount of time. Otherwise, much like an average size vs space analogy, it's convience vs security in this context.